test-data.csv contains the predictions of the test data; the file contains five columns: segment (i.e., sentiment); true_label (i.e., ground truth); id; sentiment (i.e., predicted label); probabilities (i.e., the probability for positive, neutral, and negative class).

main.js contains code for the histogram visualization (Figure 2) and code to display sentences and their predictions that are shown in the appendix (Table A IV.1, Table A IV.2).

To display the histogram and sentences, please execute:
python -m http.server